If, in the latter part of May, we take up a position at dawn in some osier bed, we listen to songs which have reached a high degree of specialisation, songs, moreover, which appeal to us on account of their beauty; if, on the other hand, we climb down the face of the sea cliff, we hear an entirely different class of songs--harsh, guttural, weird, monotonous sounds, which, appeal to us though they may, lack the music of the voices in the osier bed. And just as, in the osier bed, we can recognise each species by its voice, so we can distinguish the "cackle" of the Fulmar, the "croak" of the Guillemot, or the "grunt" of the Shag. In the osier bed, however, there is considerable variation in the song of different individuals of the same species, so much so that we can recognise this one from that; whereas on the cliff we cannot distinguish between the voices of different individuals. And the more highly developed the song, the greater the range of variation appears to be; but notwithstanding this--notwithstanding the fact that the pitch may differ, the phrase combination may differ, and the timbre may differ--the song remains nevertheless specific. So that the two principal features of "song," broadly speaking, are "diversity" and "specific character." In contrast with the call-notes, the majority of which can be heard at all times of the year, the song is restricted as a rule to one season, and that one the season of reproduction. It is true, of course, that some birds sing during the autumn, and, if the climatic conditions are favourable, in the winter also, just as others betray, in the autumn, symptoms of emotional manifestation peculiar to the spring; but just as the manifestation of the latter is feeble and vestigial, so, too, does the song of the former lack the vigour and persistency which is characteristic of the spring. Again, in contrast with the call-notes, which are common alike to both sexes, song is confined to one sex--a peculiar property of the males. Now all, I think, will agree that it must serve some biological purpose--this at least seems to be the conclusion to be drawn from the two outstanding features of "diversity" and "specific character"; and since the voices of different individuals of the same species vary, it has been suggested that, by creating a more effective pairing situation, it is serviceable in furthering the life of the individual. I do not propose at the moment to enquire whether this doctrine be true, but rather to direct attention to other ways in which the song may be useful. Is the instinct susceptible to stimulation under all conditions during the season of reproduction, or only under some well-defined condition? This is the question to which we will first direct inquiry.